A comprehensive review of elderly fall detection using wireless communication and artificial intelligence techniques

Abstract

Falls among older adults substantially affect mobility, health, and mortality. However, advancements in wireless and internet-of-things technologies have led to the development of fall detection and rescue systems. These systems offer a solution to mitigate the impact of falls by swiftly delivering emergency services to individuals, thereby reducing the risk of loss of life, injuries, and associated healthcare expenses. This paper aims to review current elderly fall detection systems (FDSs) comprehensively. The assessment examines FDSs explicitly designed for older individuals, considering fall detection methods, system architecture, wireless communications, sensor types, performance metrics, challenges, limitations, and more. In addition, a taxonomy and comprehensive review, accompanied by comparative analysis, have been conducted to categorize FDSs into traditional and artificial intelligence-based methods. The artificial intelligence techniques containing machine learning and deep learning methods for detecting elderly falls were critically reviewed and compared. Moreover, the deep learning-based systems have shown high accuracy in fall detection during the review. By conducting a comparative analysis among various FDSs, this review aims to aid researchers in identifying the most accurate and appropriate method for detecting falls in elderly individuals. In conclusion, this review assists researchers in making informed decisions and enhances the reliability and usability of elderly fall detection systems by addressing significant challenges and limitations.